Reaction before Complete Equilibrium. 67 



of the liquid only with a certain speed ; the velocity of this 

 is to be given by the equation 



~T~~ =L ^ \ t liquid ~ — * thermometer J j 

 CUT 



where t i iquid is the temperature of the liquid, t thermometer is the 

 temperature of the mercury-thermometer. Now the value of 

 C is found to be different for all mercury-thermometers ac- 

 cording as the mercury-thread is rising or falling. It is 

 considerably greater when it is rising. The above equation 

 gives no constant when t therm, becomes very near to t liq _, i. e. 

 just where such a constant must exist. For instance, we get 

 when accurate measurements of time are made in the region 

 between £n q . — £ fchei . m . = 0°'5 and t^ — t therm . = 0°* 05 for C 

 values which drop from 120° min. -1 to 20° min. -1 . As- 

 suming even that such an equation would exist down to 

 ^therm.— ^liquid =0, still we find that even for a thermometer 

 of such sensitiveness as mine C can only be less thau 



20° min.- 1 . Thus, when t^ -f^rm^To^ '!i T = °' 2 P er 



minute, i. e. the velocity with which the thermometer assumes 

 the temperature of the liquid becomes rery small even if the 

 value of C itself be very great. On the other hand, many of 

 the above reactions, such as ice-melting or ice-separation, 

 take place with such a great speed th;it the whole time of the 

 reaction, from the beginning to the end, often lasts only 

 about 15 seconds, the velocity constant of these reactions 

 being about 15 x 2 o, 3026 min." 1 . It is. therefore, quite evident 

 that the temperature indicated by the mercury 'thermometer must 

 he during the time of the reaction essentia//'/ different from the 

 actual temperature of the liquid. Now the nature of the curve 

 allows us to calculate approximately the actual temperature 

 of the liquid from those read on mercury-thermometers only 

 for the middle part of the curve which can approximately be 

 assumed to represent a straight line. At the beginning and 

 the end of the curve, on the contrary, only xery rough cor- 

 rections for ^iiq. — /th.rm. can be applied, and the absolute values 

 of these corrections prove to be so great in comparison with 

 the total value observed that we are not entitled to make any 

 use of them. Since the velocity constant varies considerably 

 with small variation in the values of t — t or t — t or . one can 

 get here almost any values one likes, so that we cannot get 

 1o the knowledge of the real nature of the phenomenon. To 

 this chief source of error we have to add some others, e. a. 

 those which arise Erom the beaker and stirrer. If these are 

 of glass or of porcelain (as it was in the case of the above 



F2 



